There’s that line about art, “good artists copy, great artists steal.” There’s some debate about if Picasso said it first, but most of us geeks know it from Steve Jobs.
Often, I see things from companies and industries outside of health care — processes, products, best practices — which inspire me. I like these little inspirations because they often aren’t rocket science, but nonetheless fuel some creative thoughts about their applicability in health care.
The other night, around 9pm on a holiday Monday, I ordered some obscure aviation stuff from Amazon. I needed a new headset, a leg-mounted chart holder, a paper calculating tool called an E6B computer and a portable canister of oxygen.
I have Amazon Prime, their subscription service which provides expedited 2 day shipping, so I expected to see my stuff on Wednesday afternoon. I was blown away when there was an Amazon box outside my door by 9am the next morning, Tuesday.
A box showed up early, big deal, right?
Here’s what I think happened and why I’m so impressed. I had been browsing for some aviation stuff for a few days. Amazon clearly knows and tracks my window shopping. It’s how they suggest items when you come back to the site.
I believe they preemptively moved some of those obscure aviation items to the closest distribution center in anticipation of my purchase. In fact, Amazon was awarded a patent for exactly that process last week.
By predicting my purchasing behavior, Amazon was able to beat my expectations for delivery — a known threat to their model is the instant gratification of local retail — and get my package to me in 12 hours.
We’ve got a lot of data in health care. That’s to the lagging but persistent implementation of electronic medical records, doctors and health systems are beginning to apply some big data science to their patient populations. For instance, any credible EMR can tell a physician how many of her patients have asthma.
More advanced systems, including bolt on solutions can look at disease panels and cross sample against last visit date. Mr. Smith, we see it’s been a year since your last visit, how’s your arthritis? Can we schedule you and appointment with Dr. Jones?
While those types of systems are starting to gain traction, the Amazon solution, despite its apparent simplicity, is far more advanced. Amazon is thinking ahead, they are predicting behavior. And with the tools we have in health care today, there’s no reason health systems and providers cannot do the same thing.
For instance, Google’s Flu Tracker looks at searches for things like flu symptoms, remedies and clinics and can accurately determine and even predict outbreaks. Providers would follow suit and move flu shots into communities before outbreaks hit. Retailers call this just in time inventory. And we don’t have to stop there.
What about actual behavior modeling? Mr. Dawson, we see from Twitter you are training for another marathon and have been skiing a lot. Studies show that preemptive sports massages can help prevent more serious injuries, can we make an appointment for you to see our physical therapist?
Yeah, ok, I secretly really want that one. But it doesn’t have to be based on leisure activities.
The point is, we have the tools and data to do some pretty impressive predictions for both populations and individuals and we’d be wise to start prototyping some of these approaches right away.
So, why aren’t we?
It’s easy to point the finger at our payment system, or internal red tape. And, I’m certain those things are a factor. But I think there’s a greater inertia at work, a sense of overwhelming change and uncertainty weighting down the industry. We’ve become cautious to the point of immobilization. If it’s not evidence based and tested by Johns Hopkins, Massachusetts General Hospital or Mayo, then we aren’t trying it.
And that’s a shame, because once Amazon figures out how to deliver a self-administering flu shot, or asthma inhaler in 12 hours, it will be too late for health care’s traditional players to catch up.
Nick Dawson is principal, better. He blogs at NickDawson.net, where this article originally appeared.